This study examines the reliability of the cardiothoracic ratio in the diagnosis of heart disease in old men. Members of the Baltimore Longitudinal Study of Aging had annual or semiannual chest x-rays as part of their health screening tests between 1958 and 1975. Measurements of cardiac diameter and of thoracic diameter were made on 1124 P-A chest x-rays from 243 individuals. Subjects ranged in age from 21 to 95 years and were categorized according to the presence or absence fo heart disease. Cross-sectional age decade means for the measured variables and for the cardiothoracic ratio were calculated. Longitudinal changes were calculated. Future plans include correlating the observed cross-sectional differences and longitudinal changes with anthropometric measurements, body mass index and blood pressure. A mortality analysis will also be performed.